Real Time Clock

Joel Sherrill joel.sherrill at
Mon May 11 13:19:28 UTC 2009

Ralf Corsepius wrote:
> Thomas Doerfler wrote:
>> Ralf,
>> if I got Sebastians proposal right, he tends to convert the existing
>> functions proprietary to the RTC drivers into "Classic API" driver
>> calls, which can then be accessed through the I/O manager API. Among
>> others benefits, this will clean up that interface (no longer a
>> proprietary function set).
>> Accessing these drivers with the POSIX standard functions
>> (open/read/write/close) will then be possible without problem.
>> Please note: we are NOT talking about SW timers, we are taking about the
>> hardware device "Real-Time-Clock".
> I am taking about API's, independently of whether they are implemented 
> in HW or SW.
>> Do you have a pointer to a proper POSIX API to access such a device?
> Off hand, I don't know if RTEMS has them, but POSIX has rtc-timer 
> functions: timer_create, timer_gettime, timer_delete, timer_settime ...
RTEMS has them and even more appropriate would
be clock_gettime() and clock_settime().  I don't know
but would think there should be a clock id which corresponds
to a real hardware clock.  I don't know if the timer_XXX()
would be right.  Would have to think.

Either way, they would have to be layered up on top of some
standard API for accessing the HW and standardizing it via
a driver interface wouldn't be bad.  How is the hw clock accessed
in GNU/Linux?
> NB: Recently, POSIX has been updated:
> POSIX-2008 is out: http:/
Without looking, is there an obvious summary of changes?
> Ralf
> _______________________________________________
> rtems-users mailing list
> rtems-users at

Joel Sherrill, Ph.D.             Director of Research & Development
joel.sherrill at        On-Line Applications Research
Ask me about RTEMS: a free RTOS  Huntsville AL 35805
   Support Available             (256) 722-9985

More information about the users mailing list